The U.S. stock market so far this month is losing ground. Should you be worried? If your answer is “yes,” you’ll probably be fixated on an indicator known as the “January Barometer,” according to which the market’s direction in January foretells its direction for the rest of the year. Through Jan. 13, the S&P 500 SPX, -0.93% is down 2.2% month-to-date. The Dow Jones Industrial Average DJIA, -1.28% hasn’t lost as much as the S&P 500, but is still down 0.1%.
My advice is to ignore the January Barometer. It is of questionable statistical significance. A good place to start this discussion is to look at what happened a year ago. U.S. stocks declined in January 2021, with the S&P 500 shedding 1.1% for the month. Far from falling from then until the end of the year, as the January Barometer predicted, the S&P 500 instead gained a stunning 28.1%.
This indicator isn’t always so off-base, but it’s been wrong more often than right in recent years. Over the past two decades, for example, it’s been right 45% of the time — less than a coin flip.
History of the January Barometer
The January Barometer was devised by Yale Hirsch in 1972. In the 1973 edition of his Stock Trader’s Almanac , Hirsch spotlighted the historical correlation between the market’s direction in January and the rest of the year. “We doubt that any technique or indicator ever devised has been so remarkably accurate as the January Barometer,” he wrote. “The barometer… has proven correct in 20 of the last 24 years… Very few stock market indicators show such an 83.3 percent accuracy for even short spans of time.”
When judging an indicator such as the January Barometer, the gold standard is how it works after it is discovered. Only if it continues to work in real-time tests can we have any confidence that it’s genuine rather than based on a spurious correlation.
To illustrate, imagine someone claims that stocks whose names begin with “G” and “T” outperform the market. As an example, over the past five years GameStop GME, -7.60% shares have gained more than 500% while Tesla TSLA, -0.84% is up over 2,300%. As I’m sure you can readily appreciate, you would be justified in giving this alleged pattern the time of day only if it continued to work over many years into the future.
The January Barometer, in fact, fails real-time tests at the 95% confidence level that statisticians often use when determining whether a pattern is genuine. Since 1972 its track record is indistinguishable from a random pattern.
Interestingly, since 1972, three months other than January do jump over this statistical hurdle: February; April, and July. Each does a better job than January of “predicting” the market’s direction over the subsequent 11 months. Imagine what this means for someone today doing the functional equivalent of what Hirsch did in 1972. Focused only on the past 50 years, this analyst would never think of creating a January Barometer, but might highlight the success of the February Barometer, the April Barometer, and the July Barometer.
None of these three Barometers would have any more legitimacy than the January Barometer did 50 years ago. My prediction: 50 years from now, it will be some other months that appear to predict the market’s direction over the subsequent 11 months.
Mark Hulbert is a regular contributor to MarketWatch. His Hulbert Ratings tracks investment newsletters that pay a flat fee to be audited. He can be reached at [email protected]